info@apippress.com
+86 18163351462 (WhatsApp)
+86 182 8014 0209
Article Content
Abstract
This paper performs a real-time forecasting exercise for US inflation from 1992Q1 to 2022Q2. We reinvestigate the literature on autoregressive (AR) inflation gap models—the deviation of inflation from long-run inflation expectations. The findings corroborate that, while simple models remain hard to beat, the multivariate extensions to the AR gap models can improve forecasting performance at short horizons. The results show that (i) forecast combination improves forecast accuracy over simpler models, (ii) aggregating survey measures, using dynamic principal components, improves forecast accuracy, and (iii) the additional information obtained from the error correction process between inflation and long-run inflation expectations can improve forecasting performance.
Notes
-
The literature also identifies market-based measures of inflation expectations as an alternative to survey data. This paper focuses on survey data of long-run inflation expectations.
-
The Livingston Survey was created in 1946 by Joseph Livingston, who was the financial editor for The Philadelphia Record. The Survey of Professional Forecasters (SPF), was initiated by the American Statistical Association (ASA) and the National Bureau of Economic Research (NBER) from 1968 to 1990. Both surveys were taken over by the Federal Reserve Bank of Philadelphia in 1990. We thank the Editor in Chief Charles Steindel for pointing this out.
-
For the explanatory variables on the right-hand side of my models, I use a mix of real-time and revised data, depending on the nature of the variable. For contemporaneous and lagged values of inflation, I use the vintage of data that would have been available at the time of the forecast. For survey expectations data, I use the original released values, as these are typically not revised.
-
Some papers have also used real-time vintage data (e.g., Kishor and Koenig 2022). Real-time vintage, on the other hand, represents the data as it was originally released or known at a specific point in time, without subsequent revisions. This vintage captures the information that was actually available to decision-makers at that moment, reflecting the uncertainties and limitations of contemporaneous data.”
-
https://www.philadelphiafed.org/surveys-and-data/real-time-data-research.
-
There is considerable debate in the literature regarding which actuals to use for economic analysis (e.g., Croushore 2010). Various actuals considered in real-time datasets include values recorded one year after the initial release, three years later, the last value before a new benchmark revision, and the latest available data. These differing actuals can potentially lead to significant variations in forecast errors. While this paper does not delve deeply into these issues, I recognize the complexities of data revisions and their potential impact on forecast evaluation.
-
Survey of Professional Forecasters (SPF), Michigan Survey of Consumer (MSC), Livingston, Blue Chip, and a constructed measure of trend inflation.
-
We use the GDPC package in R developed by Peña et al. (2020).
-
We thank the anonymous referees for their insightful comments on how the COVID-19 pandemic may affect the dynamics of inflation and inflation expectations.
